US6898254B2ExpiredUtilityA1
Turbo decoder stopping criterion improvement
Est. expiryJan 31, 2020(expired)· nominal 20-yr term from priority
H04L 1/0051H03M 13/2975H04L 1/0066
76
PatentIndex Score
27
Cited by
17
References
14
Claims
Abstract
A stopping criterion improvement for a turbo decoder that does not require division by a variable quantity. The stopping criterion improved upon generates a signal-to-noise ratio based on the mean and variance of soft-output estimates. The decoding process is aborted based on a comparison of the generated signal-to-noise ratio to a predetermined threshold.
Claims
exact text as granted — not AI-modified1. A turbo decoding system, comprising:
a decoder module, using an adaptive abort criterion to halt iteration; and
executing said adaptive abort criterion by manipulating an abort criterion equation by replacing division by a variable with multiplication by a variable thereby simplifying hardware implementation.
2. The system of claim 1 , wherein the abort criterion is based on the mean and variance of partially decoded extrinsics.
3. The system of claim 1 , wherein the abort criterion is based on a ratio of the mean and variance of partially decoded extrinsics.
4. An iterative decoder system, comprising:
a decoder module, for generating estimates of data symbols through an iterative decoding process;
a comparator having a comparison algorithm for comparing a derived quality attribute of the generated data symbol estimates to a predetermined threshold said;
comparison algorithm being executed by manipulating an abort criterion equation by replacing division by a variable with multiplication by a variable thereby simplifying hardware implementation; and
means to abort said decoding based on a comparison result.
5. The system of claim 4 , wherein the quality attribute is based on the mean and variance of the estimates of data symbols.
6. The system of claim 4 , wherein the quality attribute is based on a ratio of the mean and variance of the estimates of data symbols.
7. The system of claim 4 , wherein the quality attribute is generated without division by a variable.
8. An abort criterion in iterative decoding, comprising the steps of:
generating estimates of data symbols;
executing a comparison algorithm by manipulating an abort criterion equation by replacing division by a variable with multiplication by a variable thereby simplifying hardware implementation;
comparing the quality attribute to a predetermined threshold; and
aborting the turbo decoding based on a comparison result.
9. The method of claim 8 , wherein the quality attribute is based on the mean and variance of the estimates of data symbols.
10. The method of claim 8 , wherein the step of comparing is implemented without division by a variable.
11. An abort criterion in iterative decoding, comprising the steps of:
(a.) generating estimates of data symbols after an iteration substep;
(b.) measuring the mean of the estimates;
(c.) measuring the variance of the estimates;
(d.) generating a quality attribute based on the mean and the variance;
(e.) comparing the quality attribute to a predetermined threshold; and
(f.) aborting the turbo decoding based on a comparison result; wherein the steps (b), (c), (d), and (e) are implemented without division by a variable.
12. The method of claim 11 , wherein the quality attribute is based on a ratio of the mean and the variance of the estimates.
13. An iterative decoder system for a recursive systematic encoder, comprising:
a decoder module, wherein estimates of data symbols are generated through an iterative decoding process;
a comparison algorithm for comparing a quality attribute of the generated data symbol estimates to a predetermined threshold;
wherein the quality attribute is based on the mean and the variance of the estimates;
wherein decoding is aborted based on the result of said comparison; and
wherein the mean and variance for the estimates and the comparison algorithm are implemented without division by a variable.
14. The system of claim 13 , wherein the quality attribute is based on a ratio of the mean and variance of the estimates.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.